(Supplement) Understand Me for Life: Using Noise-augmented Automatic Speech Recognition to Improve Intelligibility in Parkinson’s Disease

Study Rationale: Most people with Parkinson’s disease (PD) develop voice and speech problems at some point during the course of the disease. Such problems, collectively known as the motor speech disorder hypokinetic dysarthria, can lead to a significant decrease in speakers’ ability to be understood and to effectively communicate with those around them. Additionally, sensory deficits are also common in people with PD, which may result in problems with voice self-perception. With this clinical trial, we will implement Understand Me for Life, a web-based app that provides noise-augmented feedback to improve speakers’ intelligibility over the course of one month of treatment.  

Hypothesis: We predict that our app will help people with PD increase their speech intelligibility, especially in noisy, everyday environments, and improve their ability to communicate with others.

Study Design: Over the course of this three-year study, an experimental group of 21 speakers with PD will self-administer, under the weekly supervision of a speech-language therapist, one month of intelligibility treatment using Understand Me for Life. The app automatically embeds the speaker’s recorded sentences in real-life noise and provides the user with a score for intelligibility (percentage of words understood in each sentence), the written sentence as recognized by the app and an audio file of the recorded sentence embedded in background noise. Another group of 21 speakers with PD will receive deferred treatment to test the app’s initial efficacy.

Impact on Diagnosis/Treatment of Parkinson’s disease: Our project represents the first speech treatment study for individuals with PD to implement feedback on speech intelligibility in noise through a participant-driven app. Our mission is to enhance individuals’ ability to self-manage and improve their intelligibility from home and to contribute to increased communicative participation and quality of life.

Next Steps for Development: Results from this clinical trial will allow our team to develop and submit a competitive NIH R01 grant to implement this AI-assisted treatment approach nationally.